Questions about Polls

Michael Barone writes a great post about the party identification on polling samples of late.

In 2004, the electorate that went to the polls or voted absentee was, according to the adjusted NEP exit poll, 37 percent Democratic and 37 percent Republican. In party identification, it was the most Republican electorate since George Gallup conducted his first random sample poll in October 1935.

But most recent national polls show Democrats with an advantage in party identification in the vicinity of 5 percent to 12 percent. Party identification usually changes slowly. Historically, voters have switched from candidates of one party to candidates of the other more readily than they have changed their party identification.

There is not currently anything posted on this phenomenon at Pollster.com, but it is an interesting idea.

If the polling sample is relating a higher percentage of self-identified Democrats, what is going on. With reputable pollsters, I have pretty good faith in their ability to draw out a solid sample, so I don't think it is necessarily the methodology, although most reputable pollsters are certainly questioning that methodology. Similarly, I don't think what we are seeing are push polls, at least not in the sense of real push polls. So what is happening?

I think the mood of the electorate is such that fewer people are willing to self-identify to total strangers that they are Republicans. Being a Republican is not generally viewed as a good thing right now and so people are inclined to say they are independent at best or Democrat at worst, even though a person calling on a poll could really care less what your party ID is, only getting throught their call list.

So are people lying? Perhaps, or perhaps the only people responding to polls are Democrats. I am not sure.

What I am sure about in regards to polling is that something smells rotten in the state of polling and I just don't know what to think. In my own state of Maryland, where a very close gubernatorial and senatorial races are taking place, you can get one set of polls from last week that says the candidates are in a dead heat, not a statistical tie, but actually tied 46-46. Then the Washington Post comes out with a poll over the weekend that says the Dems in those races hold a 10 point lead. What is to be believed? Thus, my faith in even the remote accuracy of polls is shaken.

As Barone points out:

Serious pollsters concede that there are some problems with polling. Americans have fewer landline phones than they used to, and the random digit dialing most pollsters use does not include cell-phone numbers. Larger and larger percentages of those called are declining to be interviewed.

Is polling, as a statistical science soon to undergo a transformation? I would think it needs to, in order to keep up with changes in society.